This　paper　examines　whether　journals　covered　by　the　Brandon／Hill　List　（B／H　List　journals）could　be　sorted　out　by　specific　evaluation　items　based　on　data　that　can　be　obtained　publicly．Suppose　B／H　List　journals　were　positioned　in　the　higher　rank　by　specific　evaluation　items　orcombinations　of　items，　then　B／H　List　journals　could　be　predicted　to　some　extent　using　theevaluation　items．　That　is，　it　would　enable　us　to　guess　the　selection　criteria　of　the　B／H　List　basedon　those　evaluation　items．　Although　it　would　not　be　possible　to　know　the　real　criteria，reasonable　equivalents　to　the　selection　criteria　of　the　B／H　List　could　be　attained　using　thoseevaluation　items．　　　Impact　factor，　cited　number，　citing　number，　number　of　articles，　cited　number／citingnumber　rate，　first－publication　year，　price，　price　per　issue，　number　of　libraries　archiving　thejournal　were　used　as　the　evaluation　items．　Data　concerning　these　items　were　gained　frompublic　information　resources．　Cardiovascular　system，　internal　medicine，　and　surgery　weredetermined　as　sample　fields　from　46　fields　of　the　B／H　List　2001　edition．　Journals　from　each　fieldwere　selected　from　the　journal　list　of　the　2000　edition　of　Journal　Citation　Reports　on　CD－ROMScience　edition　（JCR）．　There　were　63　cardiovascular　journals　（7　were　B／H　list　journals），　94　wereinternal　medicine　（14　were　B／H　list　journals），　and　132　were　surgery　journals　（7　were　B／H　listjournals）．　　　In　the　evaluation　items　scamined　in　this　study，　it　became　clear　that　“number　of　librariesarchiving　the　journal”　and　“impact　factor”　can　sort　out　the　B／H　List　journals　to　some　extent．However，　B／H　List　journals　were　not　completely　sorted　out　by　the　evaluation　items　used　in　thisresearch．　This　result　suggests　the　need　for　seasoning　with　the　additional　factors　such　asexperience，　the　intuition　and　other　personal　knowledge　of　the　librarian，　which　cannot　beexpressed　as　an　objective　numerical　value，　to　create　journal　selection　criteria．